Contractual obligations?

Posted Aug 18, 2015 13:17 UTC (Tue) by rsidd (subscriber, #2582)
In reply to: Contractual obligations? by dmarti
Parent article: Schaller: An Open Letter to Apache Foundation and Apache OpenOffice team

No contract as such AFAIK, but the original OpenOffice.org code was LGPL-licensed. When Oracle donated the code to Apache it chose the Apache licence (and the fact that they donated to Apache and not TDF suggests antipathy with the TDF/LO people). IBM too donated source for Symphony to Apache under the Apache licence. Both corporations stated their preference for Apache versus LGPL, so that was an early hurdle. Meanwhile, at this point LO had already been developing substantially, but taking new contributions under dual LGPL/MPL licenses. There was talk three years ago of rebasing it to the AOO apache-licensed version, going MPL-only and getting rid of LGPL, and apparently that work has progressed: today the LO webpage offers the product under the MPL but apparently some individual pieces are still LGPL'd or under other licences. But I suspect this is still the main hindrance to the Apache foundation just merging with LO -- they (AOO) don't want to switch licences.

Contractual obligations?

Posted Aug 18, 2015 13:34 UTC (Tue) by davidgerard (guest, #100304) [Link] (1 responses)

There's also who actually owns the OpenOffice.org trademark, e.g. if Apache could even give it to someone else at all.

I researched all this mess in some detail when polishing up the related Wikipedia articles - it's really confusing just what Oracle did and did not grant Apache, and the AOO project has long been actively unhelpful in making this clearer.

Searching the US trademark

Posted Aug 18, 2015 18:21 UTC (Tue) by dmarti (subscriber, #11625) [Link]

Word Mark OPENOFFICE.ORG
Filing Date March 7, 2005
Registration Number 3063339
Registration Date February 28, 2006
Owner (REGISTRANT) Team OpenOffice.org e.V. e.V. eingetragener Verein (registered association) FED REP GERMANY Team OpenOffice.org e.V. c/o Matthias Huetsch Hertogestra e 14 Hamburg FED REP GERMANY D-22111

(LAST LISTED OWNER) THE APACHE SOFTWARE FOUNDATION CORPORATION DELAWARE 1901 MUNSEY AVE. FOREST HILL MARYLAND 210502747

According to http://tmsearch.uspto.gov/ the owner is the same for the OO.o logo.

Contractual obligations?

Posted Aug 18, 2015 16:30 UTC (Tue) by rfontana (subscriber, #52677) [Link] (7 responses)

From what I understand, Oracle did not donate such code under the Apache License, but by way of an ASF 'Software Grant' agreement:
https://www.apache.org/licenses/software-grant.txt
I assume the same is true of IBM and Symphony.

Contractual obligations?

Posted Aug 18, 2015 21:27 UTC (Tue) by BrucePerens (guest, #2510) [Link] (6 responses)

Apache project developers retain their copyrights and do not assign them to the project. The project, however, doesn't attribute the developers in the code, but in a separate file. So it's impossible to tell who owns what without going through SVN checkins. And it's unsure even then.

Contractual obligations?

Posted Aug 19, 2015 3:03 UTC (Wed) by jimjag (guest, #84477) [Link] (5 responses)

That's pretty much FUD. Apache has great IP provenance.

Apache's IP provenance

Posted Aug 20, 2015 3:26 UTC (Thu) by BrucePerens (guest, #2510) [Link] (4 responses)

Not FUD at all. Read any source code file from the foundation, the provenance isn't there.

When I complained about Apache's poor IP provenance to Larry Rosen, who was the general counsel of the Apache foundation at the time, he simply didn't believe that maintaining the provenance of individual authors with the portion of the code that they contributed was important. What has changed since then?

Apache's IP provenance

Posted Aug 20, 2015 12:59 UTC (Thu) by jimjag (guest, #84477) [Link]

The provenance is in the record keeping of grants, iCLAs and the SVN logs. It's pretty in-depth and the model for how many, many other FOSS projects handle IP provenance.

Apache's IP provenance

Posted Aug 20, 2015 13:23 UTC (Thu) by jimjag (guest, #84477) [Link] (2 responses)

Larry Rosen was never General Counsel of the ASF. And despite Larry's opinion (many of which I respect), we have numerous legal opinions to the contrary.

Apache's IP provenance

Posted Aug 25, 2015 9:07 UTC (Tue) by edomaur (subscriber, #14520) [Link] (1 responses)

He was a member of the Board of Directors for some time in 2011 and resigned the same year. Perhaps that was what Bruce Perens refers to.

Apache's IP provenance

Posted Aug 25, 2015 17:06 UTC (Tue) by jimjag (guest, #84477) [Link]

Possibly... he did say General Counsel though, which is quite different from Director. But it may have been a typo.

Contractual obligations?

Posted Aug 19, 2015 15:34 UTC (Wed) by jimjag (guest, #84477) [Link] (8 responses)

At the time that Oracle was looking for a place to donate OO, TDF did not exist. There was no legal entity to donate it to. Plus, as you state, there was a desire to have OO under ALv2. So really, with there being no legal entity to accept, and the then-LO community not wanting an ALv2 versions office suite, there was really no other place to go. The hope was that by donating to the ASF, that AOO would be able to form an ALv2 licensed core office framework that LO and the entire OO eco-system could use, consume and leverage as needed.

As you note, there were efforts to combine forces, but it always boiled down to ALv2 versus copyleft. As much as they (AOO) don't want to switch licenses (to a copyleft), they (LO) also don't want to switch licenses (to ALv2, or some other permissive one). It is useful to recall that one of the primary ways that LO was able to rebase and relicense their code was by basically consuming AOO and then relicensing THAT under MPL. It was only because of the software grant of OO to Apache, and the subsequent relicensing of that to ALv2, that LO was able to relicense itself, being based and forked from the original OO code. Placing full "blame" (for lack of a better term) under AOO is really disingenuous and quite unfair.

I find the whole situation both sad and ironic. After all, the whole point of copyleft licenses is basically to force someone to do something which they should do anyway, as a moral imperative (basically, in the case of weak copyleft, "give back" changes/fixes/patches of a consumed codebase back to the upstream community). It's basically to enforce a two-way street. Yet the whole AOO/LO debacle has been basically one-way. "Yeah, well, that's what you get when you license under the ALv2"... true. But that kind of misses the whole point, doesn't it. Choosing a license which forces others to do what you yourself won't.

Oracle and Apache

Posted Aug 20, 2015 3:54 UTC (Thu) by BrucePerens (guest, #2510) [Link] (2 responses)

Although the Document Foundation might not have existed, the LibreOffice project did, and there were any number of proper 501(c)3 organizations that would have accepted code for the LibreOffice project and under their control. SFC, SPI, Mozilla Foundation, were all around. And then there were not-for-profits that were not 501(c)3s, like Linux Foundation, The Open Group, and probably another 100 organizations that we could have found in the community.

It also would have been trivial for Oracle to found a legal entity to hold the code, to their custom requirements.

So, yes, LO was able to use the code. But not only was the contributor not assisting LO, the contributor started a fork within Apache which was positioned against LO from day one, and had its own anti-LO PR spokesperson whom I think IBM was paying for and tolerated for a good long time.

So if you're gonna call folks disingenuous and unfair, I hope you can take what you dish out, because positioning AOO as any sort of attempt to help LO really sounds disingenuous and unfair. From here, it looks like LO has succeeded dispite the Apache Foundation and its partners Oracle and IBM.

The rest of your comment is just venting your dislike of the GPL. And it's disingenuous to accuse GPL participants of not giving back because they used the GPL, when they might have been more willing to use the Apache license had you not presented them with an undesired fork and a great deal of hostility.

From the outside, Apache seems to have developed some negative characteristics. GPL-hating worse than was ever expressed by Brett Glass when he was a BSD OS supporter. Willing to work against its own community.

And yes, you don't see yourselves this way and thus it seems ironic to you.

Oracle and Apache

Posted Aug 20, 2015 13:17 UTC (Thu) by jimjag (guest, #84477) [Link]

It also would have been trivial for Oracle to found a legal entity to hold the code, to their custom requirements.

They did. It was the ASF.

They wanted a legal entity to "have" the code, as well as for it to be under an ALv2 permissive type license.

Are you honestly saying that if, for example, the SFC would have accepted the code, and it would have then been under ALv2 there as well, that things would be "different"???? Or are you suggesting that Oracle should have donated it to SFC, for example, to simply "hold" until TDF was legal? Well, maybe, but they didn't. That is hardly Apache's fault.

And let's not forget, if Apache had not accepted the code, and had not then been able to relicense the entire suite to ALv2, then LO would not have been able to relicense their code to what I assume is a Good Thing for them.

As far as "disliking" the GPL, well, I submit that TDF/LO also "dislike" the ALv2 (and yet, Apache is being painted as the intolerant one... interesting). But even though I don't dislike the GPL (it's not my favored license, yet I code quite a bit of GPL code and have no issues supporting said projects), it is clear that for you and others, it's all about not seeing the need or desire for a permissively licensed OO suite or framework. That's fine.

Oracle and Apache

Posted Aug 20, 2015 13:37 UTC (Thu) by jimjag (guest, #84477) [Link]

And it's disingenuous to accuse GPL participants of not giving back because they used the GPL, when they might have been more willing to use the Apache license had you not presented them with an undesired fork and a great deal of hostility.

The hostility was there way before Apache got involved.

I submit that if TDF had existed as a legal entity and if they had been willing to have LO under ALv2 (or some other permissive license) that Oracle would have donated it to them.

I will go further. If the conditions of the above had been true, and Oracle had not donated it to the TDF, then the ASF would not have accepted it. I and others would have strongly rejected any such proposal.

Contractual obligations?

Posted Aug 20, 2015 4:45 UTC (Thu) by rsidd (subscriber, #2582) [Link] (4 responses)

At the time that Oracle was looking for a place to donate OO, TDF did not exist.

Odd that you would say that shortly after linking to mails from the time on a list called tdf-discuss. Yes, the legalities weren't yet done, but TDF very much existed.

Contractual obligations?

Posted Aug 20, 2015 13:03 UTC (Thu) by jimjag (guest, #84477) [Link] (3 responses)

I wrote:

"TDF did not exist. There was no legal entity to donate it to."

Which is true and was the whole point. Yet you conveniently chose to note quote *the exact next sentence* which ties it together. If the TDF had existed, as a *legal entity*, then it is possible that Oracle could have donated OO to it. But Oracle also wanted OO to be under a permissive ALv2 (or similar) license.

Contractual obligations?

Posted Aug 20, 2015 17:24 UTC (Thu) by rsidd (subscriber, #2582) [Link] (2 responses)

In all your arguments above, you skirt one point: LibreOffice had been developing the OOo codebase further, had one release six months previous to the incubation suggestion, and even before LO existed, go-oo.org had been making contributions (merged into LO but not into OOo or AOO) that had been adopted by most Linux distros. The incubator proposal made no mention of any of this. What was the proposal? To throw all this in the trashcan? To pretend it never happened? Because there is no mention, in anything you linked, on how to incorporate it into AOO. If there was no such plan, who can imagine the incubator proposal was in good faith? Whether TDF had been legally incorporated or not is a red herring. The codebase had been moving along without Sun's/Oracle's help for a while already.

Contractual obligations?

Posted Aug 25, 2015 17:07 UTC (Tue) by jimjag (guest, #84477) [Link] (1 responses)

Not a red herring but a substantial fact to the whole issue.

Contractual obligations?

Posted Aug 25, 2015 17:15 UTC (Tue) by rsidd (subscriber, #2582) [Link]

Does one read that as saying the proposal was to trash all go-oo/LO development that had occurred up until that point? If so, thanks for (sort of) clearing that up. If that's not what you're saying, why are you continuing to avoid that question?